The current body of knowledge regarding the etiology and prevention of cancer draws substantially from prospective epidemiologic studies, among which African-Americans have been greatly underrepresented. Meanwhile, African-Americans experience a disproportionate incidence of and/or mortality from many major cancers and other chronic diseases for reasons that remain unknown. Our objective is to initiate a long-term prospective cohort study comprised of 90,000 African-American and 45,000 non African-American residents of southeastern US states. Comprehensive baseline information will be collected about dietary, lifestyle, medical, occupational and other factors, and a large biospecimen repository will be established that can be used to test future hypotheses involving individual susceptibility to environmental carcinogens. Seventy thousand of the cohort members will be recruited from southern community health centers, providing health services to primarily low-income residents without health insurance. Here participants will be offered an in-person assisted interview using a structured questionnaire and the collection of buccal cell and blood specimens in a practical and convenient setting. The remaining 65,000 will be recruited from the general populations (identified from drivers license, voter registration and Medicare files) of the southern states via mail and telephone contact. For these individuals a portion of the interview will be conducted by phone, with the remainder of the questionnaire (mostly the dietary section) mailed to be self administered. The general population sample will also be mailed a buccal cell collection kit and return mail envelope. The cohort will then be followed actively via periodic repeat contact to ascertain health status and exposure changes since entry, and passively via linkage with state mortality and cancer registries and the National Death Index. The cohort was designed to be large enough to begin analyses of specific risk factors for common cancers (prostate, lung, breast, colorectal) within the 5-year study period. Furthermore, the large study size will enable the assessment of less common cancers afflicting blacks more than other racial groups shortly thereafter, and the biospecimen repository will provide an invaluable resource for the evaluation of biomarkers of cancer risk. Pilot testing in Tennessee, Mississippi and Florida has indicated that the proposed research approach is feasible. The resultant study should help answer questions regarding the etiology of certain cancers, elucidate causes of the disparities in cancer incidence and mortality across racial groups, and lead to the development of measures aimed at the prevention of cancer and other diseases, especially among African-Americans.